Outsourcing Cognitive Control to the Environment: Adult Age Differences in the Use of Task Cues

Article excerpt

When an initial phase of cued task switching is followed by a phase of single-task trials, older adults show difficulties changing to the more efficient single-task mode of processing (Mayr & Liebscher, 2001). In Experiment 1, we show that these costs follow older adults' continued tendency to inspect task cues even though these provide no new information. In Experiment 2, we included a condition in which task cues were eliminated from the display after the task-switching phase. In this condition, older adults behaved the same as younger adults, suggesting that the presence of the task cue is critical for observing age differences while switching from a "high-control" to a "low-control" mode of processing. We discuss our results in terms of a life-span shift with regard to the reliance on internal versus external sources of information under conditions of high-control demands.

Assume you are about to prepare a meal for friends coming over for dinner. Initially, you will spend some time thinking about necessary ingredients and the exact sequence of steps. This phase requires deliberation and active processing of external information (e.g., from cookbooks). Once the plan is laid out, and assuming everything goes as planned, you should be able to execute the remaining routine activities in a relatively ballistic manner, freeing up mental capacity for other activities such as chatting with your arriving friends. Many day-to-day situations require such a back and forth between high-control and low-control phases.

Ideally, we should enter a low-control mode whenever possible and use the costly high-control mode only when necessary. Are we able to rapidly switch from high-control phases requiring intense, deliberate information processing, to phases in which information intake is geared toward the demands of routine activities?

Normal younger adults appear to have little difficulty in shifting control modes. Using a variant of the task-switching paradigm, Mayr and Liebscher (2001) required younger and older participants to switch between responding to the color or form of a stimulus on the basis of task cues (color and shape) presented above the stimulus. On the first 40 trials, individuals switched randomly from one task to another, prompted by the cue. This situation requires significant top-down control and active use of the cue to select the relevant task set on each trial. However, from Trial 41 onward, one of the two tasks dropped out, and only one task was relevant for the remaining 80 trials. Individuals were informed which task would drop out at the start of the block, and the no-longer-relevant task indicator was "marked out" to emphasize that this task was now irrelevant. Obviously, during this so-called fade-out phase, top-down task selection and task cue use are no longer necessary, and individuals should switch into a low-control mode. Indeed, only 10 trials into the fade-out phase, the younger adults' performance was identical to their performance in a single-task control situation in which only one task was relevant throughout the block. In contrast, older adults showed a large, 300-msec "fade-out" cost (fade-out phase RT minus corresponding single-task RT) after one of the two tasks dropped out, and a 100-msec cost remained at the end of the block. In other words, older adults continued to experience a significant influence of the initial task-selection phase long after younger adults had ceased to show any such influence. These results suggest that the selection of control modes is age sensitive, and that a better understanding of the general problem of control mode selection should result from a closer examination of how older and younger adults transition between control modes.

One potential reason for older adults' failure to transition fully to single-task performance levels is suggested by studies of visual perception (see, e.g., Ballard, Hayhoe, Pook, & Rao, 1997) that indicate that people encode limited information about the environment, preferring to use the environment as an external memory. …